Chapter 12 - Cartography and map production

We begin a new section of the book, turning from organizational and global issues to the analytical problems that underly GIScience, topics that are central to the kind of work that we do as environmental scientists and managers. 

In my opinion cartography is a broad and critical area of the data → knowledge process that also has deep esthetic roots. As you read this chapter, see if you can count the number of examples that have some environmental theme. I can only find a few.

12.1 Introduction

This new section begins with a diverse "gallery" of map styles that nicely illustrate the "craft" of cartography, between art and science.

Figures 12.1 and 12.2 are a juxtaposition of maps from two national mapping organizations that supposedly have different users. But as a USGS scientist who has worked with NOAA - and as a kayaker whose realm is the land-water interface - I'm frustrated by the co-existence of very different kinds of maps (NOAA calls them "charts" and USGS calls them "quads'!) that are difficult to integrate. Here for example is an conflation of a USGS and a NOAA chart for the US National Mall. Interestingly, a 2008 article in Science magazine suggested merging the two agencies, but I'm told not to hold my breath, as this is an ancient schism. (Even if you're not interested in the bureaucratic issues, skim the article and study the graphic on page 2.)

One of our themes is GIS as integration of attributes (Phenomena) and locations (Space). Figure 12.3B adds statistical information to the map, something that ArcGIS now facilitates. I usually do this outside of the GIS then arrange statistical graphics along with the map. Note that the map has Darfur floating in empty space.

12.2 Maps and cartography

We needn't get hung up on the paper v. digital map dichotomy. You could easily fill a page - or a book -  with the virtues and shortcomings of each. In fact, a simple exercise would be to transfer the discussion of this Section 12.2.1 into such a table:
ADVANTAGES DISADVANTAGES
PAPER
DIGITAL
My favorite plus for paper is that no electricity is required.

All of the maps in this book became digital at some point, but you should study a few for what kind of information (and misinformation?) they convey. For example:
    PLUS: The Australia Boating Safety Chart (Figure 12.2) could be a trusty companion navigating the obviously tricky waters of Tin Can Inlet, but
    DELTA: at first glance the Southern US Population Density map Figure 12.5B tells me that Miami is an area of low density because someone forgot to "Flip Symbols" (as we say). So even though this class introduces you to the high-tech world of GIS, you're also becoming a more discerning map consumer.

Note that the two ESRI maps in this section are based on templates that can be found at
C:\arcgis\Bin\Templates\USA\SouthernUSA.mxt
so you can immediately make a map from the data therein, or add your own data into a cartographically "correct" layout. To do this, run ArcGIS then browse to the templates.

12.3 Principles of map design

At the moment I'm reading Edward Tufte's latest book Beautiful Evidence, the fourth in his series preaching the canons of information architecture. Each of his books - which I strongly urge you to look at - features ideas about mapping.

This section reminds us to be clear on the matter of map scale, and refers back to Box 4.2. Look at the index for all the references to this tricky subject.

The ESRI templates are a useful guide to the components of a map, as is Figure 12.8. If you want a quick guide to these elements run ArcGIS > Layout View (the second tab at the bottom) > Insert then see a list of objects you can add, in (my opinion of) order of importance being scale, legend, and title. (I think that the book's example has too many Biome Types at similar colors to be displayed at a scale of 400,000,000.)

Tufte shares the pinnacle of info architecture with Jacques Bertin, whose work has been summarized by many, including geographers, as in Figure 12.9. You may not agree with this matrix, but it certainly makes you think about geometry and symbology. It is interesting to ponder the fact that none of the types of geometric objects represented is actually visible (even a polygon has zero thickness), only when we represent them do they become part of the map. You might want to reverse the order of the columns to conform to our D = 0, 1, 2 convention. Also, rows 2-4 are a convenient reference to the confusing subject of "color."

This argument is taken to a practical level in Table 12.1; a digital version of this book would have links from the cells of the table to the dozen examples cited. From an analytical point of view lines are much less important that point measurements and polygon values. Rater data should also be added here.

We saw in GTKAGIS Chapter 6 (and especially Exercise 6a) how classification of real-valued data leads to radically different impressions of spatial distribution, as further illustrated in Figure 12.13: same data, different information. On a technical note, I would reverse the ordering of the colors and the labels, showing the more "intense" values at the top of the legend; and I'd show the surrounding states and country in a light gray, etc. Some people don't know the US Upper Midwest... This figure should be compared to Figure 4.10, as well as my own choropleth experiment. (And if you're interested to know more about symbology, check out Cindy Brewer's website.)

Here are the breaks for the 3 Mobile Home Density maps, as well as my preferred precision:


BREAK		 NATURAL
BREAKS		 EQUAL
INTERVAL
QUANTILE
4 11.9   20.2   5.4  
3 6.8   15.8   3.5  
2 4.1   10.2   2.4  
1 2.1   5.1   1.5  

And see as well the histograms I made of the different data. Standard deviation classification highlights extreme values.

This section justifiably argues that cartography is a kind of statistical analysis, but GIS programs are just giving us features that facilitate this (see Figure 12.14.8). Another leader in this area is Dan Carr; you can find some of his maps beginning at (.pdf) page 6 in a 1998 newsletter.

12.4 Map series

This section explores in greater depth the idea of a mapping "tradition" that inheres in various organizations, and therefore explains e.g. the NOAA/USGS divergence I was bemoaning above. About 10 years ago the USGS National Mapping Division completed its 55,000th and final 8½ minute 1:24,000 scale paper topographic quadrangle, and a great project came to an end. Whether the National Map & Atlas will top this achievement remains to be seen.

The great advantage of a map series is that once the user becomes familiar with the consistent cartographic style they don't have to interpret what they see: objects on maps at the same scale can be compared, the same area is covered, colors and symbols have the same meaning, etc. As we are experimenting with GIS, your assignments and projects won't achieve this consistency, but it is a laudable goal.

12.5 Applications

Longley et al are constrained to static examples, yet the latest cartographic frontiers integrate time/space/phenomenon to provide animation and the analysis of interaction. It is clear from this section that the paper maps on these pages can only begin to suggest the rich interrelationships that we can explore with GIS, but if you're trying to produce a graphic for a website or to put on a piece of paper, this is an excellent reference chapter.

Yet, within the confines of a static display, a multi-frame map like the pipeline system in Figure 12.20 can show multiple representations of the same phenomenon, in this case the locations of the pipes and trees showing hierarchy and valves. The staff of the mapping agency spent many hours developing this scheme so that it can be readily used to visualize different parts of the system. And the user can print it out and take it on a helicopter!

I expect that the physical - and even the paper - map will be around for a long time, but even though I have never made a finished cartographic product, I can use GIS to create maps nearly the equal of the finest craftsperson in their informativeness - if not their esthetics! And you can too.